Crossword

12345678910111213
Across
  1. 5. Set of all tokens
  2. 6. Start, end, etc.
  3. 7. Splits text into tokens
  4. 10. Commas, periods, etc.
  5. 11. Numerical representations of text
  6. 13. Spaces and tabs
Down
  1. 1. Tunable parameter
  2. 2. Compressing long prompts
  3. 3. Reducing sequence length
  4. 4. How often pairs appear
  5. 8. Byte Pair Encoding
  6. 9. Handling various data types
  7. 12. Large Language Models